Zebrafish of the Tübingen strain were maintained in the zebrafish facility of the Institute of Feed Research, Chinese Academy of Agricultural Sciences (Beijing, China). The size of each tank was 25.5 cm × 18.5 cm × 18.0 cm. During the 4-week feeding period, the water in the rearing system was kept running, the rearing temperature was 25–28°C, the dissolved oxygen was >6.0 mg/L, the pH was 7.0–7.2, the nitrogen content was <0.50 mg/L, and the nitrogen content (as NO₂) was <0.02 mg/L. Zebrafish were maintained at a 14:10 L:D cycle.

The feed formulation is presented in Supplementary Table 1. Casein, soybean oil, and wheat flour were used as dietary protein, lipid, and carbohydrate sources, respectively. The control-check diet (CK) was supplemented with soybean oil at 60 g/kg (Supplementary Table 1). Increased levels of disodium succinate (0.05, 0.1, 0.15, and 0.2%) were compensated by decreasing equal levels of zeolite powder to prepare 0.05, 0.10, 0.15, and 0.2% succinate-supplemented diets (S0.05, S0.10, S0.15, and S0.2). All dry ingredients were ground through a 60-mesh screen. Diets were prepared by manually mixing the dry ingredients with oil and water. Each diet was extruded in a manual extruder with a 2.5-mm aperture. The pellets were air-dried and stored at −20°C in plastic bags in small quantities and were ground through a 30-mesh screen prior feeding. During the feeding trial, healthy, uniformly sized 1-month-old zebrafish (0.861 ± 0.002 g/20 fish) were randomly divided into five groups and fed the CK, S0.05, S0.10, S0.15, or S0.2 diet (Supplementary Table 1). Zebrafish were fed to apparent satiation two times daily at 9:00 and 16:00. The feed amount for each group was recorded accordingly.

All fish were anesthetized with tricaine methanesulfonate (MS222) before sampling. At the end of the 4-week feeding trial, the survival rate in each group was recorded (100 × final survival individuals/initial individuals); the fish in each tank were weighed to calculate weight gain [100 × (final body weight−initial body weight)/initial body weight] and feed efficiency [100 × (final body weight−initial body weight)/feed intake]. Whole fish were collected for analysis of whole body proximate composition and whole body energy gain. Intestine, liver, and muscle samples were collected for the detection of metabolic intermediates, gene expression, and protein succinylation. Intestine samples were isolated for the detection of digestive enzyme activities. Randomly selected living zebrafish from each group were used to evaluate standard metabolic energy, feeding metabolic energy, fasting plasma glucose, and glucose tolerance.

After the 4-week feeding trial, zebrafish were anesthetized with MS222. After removal of water from the body surface, zebrafish were weighed and frozen in liquid nitrogen. Body proximal composition was detected as previously described (24, 25). Moisture contents were determined by an oven drying method. Whole body protein contents were determined by a semiautomatic Kjeldahl system. Whole body fat contents were determined using the Soxtec method with petroleum ether as the solvent. Ash contents were determined by a high-temperature burning method.

The energy values of fish body were measured with a calorimeter. Initial fish samples were collected at the beginning of feeding. Final fish samples were collected at the end of a 4-week feeding trial. All samples were frozen at −80°C and subject to freezing-drying. The dry weight of each sample was recorded. Energy gain (%) = (final energy value−initial energy value)/initial energy value × 100%.

Standard and feeding metabolic energy were assessed by oxygen consumption (OC) monitored in an 8-chamber (2.3 L) respirometer (15, 25). Before being transferred to the chambers, zebrafish were fasted for 24 h. An empty chamber was used as a blank control. Then, zebrafish were fed equal amount of CK or S0.15 diet for 3 days. On the 4th day, the water in the chambers was changed and initial oxygen contents in water were measured. After the transfer, the fish were fed for one more day. Final oxygen contents in water were measured after 24 h (the 5th day). OC measured during the 24-h period reflected total metabolism. Then, the zebrafish were fasted for the following 3 days (the 5th to 7th day). On the 8th day, the water in chambers was changed and initial oxygen contents in the water were measured. Then, the fish were fasted for another day. Final oxygen contents in the water were measured after 24 h (the 9th day). OC measured during the 24-h period reflected standard metabolism. Oxygen was monitored using a polarographic oxygen electrode with a temperature sensor (lower limit: 4.0 mg/L, Yellow Springs Instruments, Inc.). OC was converted to energy by using a conversion factor of 13.84 J/mg O₂. Feeding metabolic energy (J/day) = total metabolic energy − standard metabolic energy.

Intestine samples were collected 4 h after the last feeding and homogenized in sterile PBS (wt:vol = 1:9). Homogenates were centrifuged at 13,201 g at 4°C for 10 min, and supernatants were collected for the detection of digestive enzyme activities. Enzyme activities were analyzed with the SynergyMX Multi-Functional MPP Detector (Biotek, United States) as previously described (15). Digestive enzyme activities were expressed as activity unit per gram protein (U/g protein).

The swimming activities of zebrafish larvae fed CK or S0.15 diet (Supplementary Table 2) were measured for 1 week using an infrared (IR) tracking device (WMicroTracker; 15, 26). Zebrafish larvae were distributed into a 24-well microplate at a density of 1 larva per 2 ml medium. The swimming activities of the larvae were monitored for 72 h. Zebrafish larvae were fed two times daily (9:00, 17:00). Swimming activities were calculated by the cumulative times of activity events happened in every 30 min.

Intestine, liver and muscle samples were collected for the colorimetric or fluorescent measurement of succinate, pyruvate or acetyl-CoA (Sigma-Aldrich, United States) using 96-well plate and with the SynergyMX Multi-Functional MPP Detector (Biotek, United States). Standards for succinate or pyruvate (0, 2, 4, 6, 8, and 10 μl of 1 mM solutions) were pipetted into standard wells generating a concentration gradient of 0, 2, 4, 6, 8, and 10 nmol/well. Succinate or pyruvate assay buffer was then added to each standard well to bring the total volume to 50 μl. Tissues were rapidly homogenized in 100 μl of ice-cold succinate or pyruvate assay buffer and subjected to centrifugation at 10,000 g for 5 min to remove insoluble materials. Supernatants (50 μl) were mixed with 50 μl of the reaction solution and incubated for 30 min at 37°C in the dark. Absorbances were measured at 450 nm for succinate and 570 nm for pyruvate. Concentrations were calculated according to the standard curve and expressed as nanomole per milligram tissue protein (nmol/mg protein).

Standards for fluorescent detection of acetyl-CoA (0, 10, 20, 30, 40, and 50 μl of 2 μM standard solution) were placed into a black 96-well plate, generating concentrations of 0, 20, 40, 60, 80, and 100 pmol/well. Acetyl-CoA assay buffer was then added to each standard well to bring the total volume to 50 μl. Tissues were rapidly homogenized in 100 μl of ice-cold acetyl-CoA assay buffer and subjected to centrifugation at 10,000 g for 10 min to remove insoluble materials. The supernatants (50 μl) were mixed with 50 μl of reaction solution and incubated for 10 min at 37°C in the dark. Fluorescences were measured using an extinction/emission wavelength of 535/587 nm. The level of acetyl-CoA was expressed as nanomole per milligram tissue protein (nmol/mg protein).

Tissue TG was extracted by chloroform: methanol (2:1). Fresh tissues were homogenized in 1 ml of PBS and then vortexed with 5 ml of chloroform: methanol (2:1) vigorously. After standing for 10 min, the TG-containing bottom-layer was extracted and solvent was evaporated at 70°C under nitrogen. Then, 1% Triton × 100 (prepared in chloroform) was added to solubilize the TG. The mixture was then heated at 70°C and again dried under nitrogen following the addition of deionized water to disperse the TG. The TG content was determined by a quantitative enzymatic method as described in a previous study (27) and was expressed as TG weight per gram tissue (mg/g tissue).

Zebrafish were fasted for 72 h. Groups of zebrafish were then used to collect blood using a standard sampling protocol (28). Briefly, blood samples were collected in sodium heparin-coated tubes from the zebrafish following cutting off the tail. Plasma was then produced by centrifugation (1,467 g/min for 10 min). The remaining zebrafish were then intraperitoneally injected with glucose [1 g/kg body weight (b.w.)] for the glucose tolerance test. Blood was collected after 1, 2, 4, 5, and 6 h. Plasma glucose was measured using the Glucose Assay kit (Nanjing Jiancheng Bioengineering Institute, China) according to the manufacturer’s instructions. Plasma glucose was expressed as mmol of glucose per liter plasma (mmol/L).

Total RNA extraction, reverse transcription, and qPCR were conducted as described previously (29). Total RNA was isolated using the Trizol reagent (Cwbio, Beijing, China) and then reversed transcribed to cDNA by FastKing gDNA Dispelling RT SuperMix (Tiangen, Beijing, China). The qPCR was performed using SYBR^®Green Supermix according to the manufacturer’s instructions (Tiangen, Beijing, China). The results were stored, managed, and analyzed using LightCycler 480 software (Roche, Basel, Switzerland). The qPCR primers used are listed in Supplementary Table 3.

The assay of protein succinylation was performed by Jingjie PTM BIO (Hanzhou, China). Intestine and liver samples were grinded by liquid nitrogen to cell powder and then lysed with lysis buffer (3 μM trichostation and 50 mM nicotinamide), followed by centrifugation at 12,000 g at 4°C for 10 min to remove debris. Then, the supernatants were collected and protein concentrations were determined with the BCA kit according to the manufacturer’s instructions (Sigma, United States). To enrich the modified peptides, tryptic peptides dissolved in NETN buffer (100 mM NaCl, 1 mM EDTA, 50 mM Tris–HCl, 0.5% NP-40, and pH 8.0) were incubated with pre-washed antibody beads (PTM402, PTM Bio) at 4°C overnight with gentle shaking. Then, the beads were washed four times with NETN buffer and two times with H₂O. The bound peptides were eluted from the beads with 0.1% trifluoroacetic acid.

For liquid chromatography with tandem mass spectrometry (LC-MS/MS) analysis, the resulting peptides were desalted with C18 ZipTips (Millipore). The tryptic peptides were dissolved in 0.1% formic acid (solvent A), directly loaded onto a home-made reversed-phase analytical column (15 cm length, 75 μm i.d.). The gradient comprised an increase from 6 to 22% solvent B (0.1% formic acid in acetonitrile) over 43 min, 22 to 30% in 13 min, and climbing to 80% in 2 min and then holding at 80% for the last 2 min, all at a constant flow rate of 450 nl/min in an NanoElute UPLC system. The peptides were subjected to a capillary ion source for ionization and analyzed by TIMS-TOF Pro mass spectrometry. The electrospray voltage applied was 1.6 kV. The m/z scan range of the secondary mass spectrometry was 100 to 1,700. The data acquisition mode is parallel cumulative serial fragmentation mode. The resulting MS/MS data were processed with the Maxquant search engine (v.1.6.6.0). Tandem mass spectra were searched against Danio_rerio_7955_PR_20191112 uniprot database concatenated with the reverse decoy database. Trypsin/P was specified as the cleavage enzyme, allowing up to two missed cleavages. The mass tolerance for precursor ions was set to 20 and 20 ppm in the First search and in the Main search, and the mass tolerance for fragment ions was set to 0.02 Da. Carbamidomethyl on Cys was specified as a fixed modification, and acetylation modification and oxidation on Met were specified as variable modifications. False discovery rate (FDR) was adjusted to <1%.

Gut contents were collected at 4–6 h after the last feeding and were used to analyze gut microbiota composition. The 16s V3–V4 region was amplified using primer pairs 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′). 16S ribosomal RNA gene sequencing was performed by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China) using the Illumina MiSeq PE300 platform (Illumina, San Diego, United States). Then, the raw pair-end readings were subjected to a quality-control procedure using the UPARSE-operational taxonomic unit (OTU) algorithm (30). The qualified reads were clustered to generate OTUs at the 97% similarity level using the USEARCH sequence analysis tool (30). A representative sequence from each OTU was assigned to a taxonomic level in the Ribosomal Database Project (RDP) database using the RDP classifier (31).

Statistical analyses were conducted using GraphPad Prism 5 software (GraphPad Software, Inc., San Diego, CA, United States). The results are expressed as means ± standard errors of the means (SEMs). Comparisons between two groups were analyzed using the Student’s t-test, and comparisons between multiple groups were analyzed by one-way analysis of variance followed by Duncan’s test. Statistical significance was set at p < 0.05.

After the 4-week feeding trial, no mortality was observed in any of the groups (Figure 1A). Daily feeding rates were increased significantly (p < 0.05) with 0.05, 0.1, or 0.15% succinate supplementation, but non-significantly increased with the additive amount of 0.2%, suggesting an appetite-promoting effect of succinate (Figure 1B). Body weight tended to increase with inclusion (0.05 and 0.1%) up to 0.15%, which was significantly higher than that in fish fed the CK diet (105.7 vs. 90.9%, p < 0.05; Figure 1C). There were no differences in weight gain between the S0.2 and the other groups (Figure 1C). Feed efficiency was the same among all groups (Figure 1D). Thus, 0.15% succinate was used as the optimal dose in the following experiments.

The chemical composition of the whole body was analyzed after a 4-week feeding trial. Moisture and crude ash contents were similar between the CK and S0.15 group (Figures 2A,B), while crude lipid (23.1% vs. 28.7%, p < 0.01) and protein contents (14.74 vs. 16.52%, p < 0.05) were significantly increased in fish fed the S0.15 diet (Figures 2C,D).

Accordingly, zebrafish fed the S0.15 diet had significantly higher energy gain than those fed the CK diet (23.8 vs. 21.8%, p < 0.05; Figure 3A). Standard metabolic energy in the S0.15 group was 40.3% lower than that in the CK group (p < 0.01; Figure 3B). However, feeding metabolic energy (Figure 3C), digestive enzyme activities (lipase, protease, and amylase; Figure 3D), and swimming activities (Figures 3E,F) were similar between the two groups. These results indicate that dietary succinate could promote fat, protein, and energy deposition in the whole body of zebrafish.

Compared with the CK group, the S0.15 group showed a significantly increased level of succinate in the intestine (2.2-fold, p < 0.05), but a reduced level in the liver (0.66-fold, p < 0.05), with no change in the muscle (Figure 4A). The level of pyruvate was significantly increased in the intestine (1.3-fold, p < 0.01) of the S0.15 group, with no differences in both the liver and muscle compared with the CK group (Figure 4B). The levels of acetyl-CoA were significantly elevated in the intestine (2.3-fold, p < 0.05) and liver (2.2-fold, p < 0.05) of the S0.15 group, with no changes in the muscle (Figure 4C). These results suggest that dietary succinate regulates the levels of metabolic intermediates in the intestine and liver.

Zebrafish fed the S0.15 diet had 54.5 and 96.8 mg of TG per gram tissue in the intestine and liver, respectively, (Figure 5A). This was significantly higher than that in the CK group, which had only 32.1 and 32.4 mg/g TG in the same tissues, respectively (p < 0.05; Figure 5A). Feeding the S0.15 diet apparently increased hepatic fat contents in the intestine and liver, as shown by H&E and oil red O staining (Supplementary Figures 1A–D). However, no significant changes in muscle TG content were observed in the S0.15 group (Figure 5A and Supplementary Figure 1E). The basal plasma glucose level in zebrafish fed the S0.15 diet was significantly lower than those fed the CK diet (1.9 vs. 2.6 mM, p < 0.05; Figure 5B). Moreover, feeding the S0.15 diet significantly improved glucose tolerance, as shown by the significantly higher rates of glucose clearance when monitored at 2–6 h after intraperitoneal injection (p < 0.01; Figure 5C). These results demonstrate that dietary succinate could promote hepatic and intestinal fat accumulation while improve glucose homeostasis.

The expression levels of intestinal lipid synthesis-related genes acetyl-CoA carboxylase 1 (acc1; 1.8-fold) and fatty acid synthase (fas; 1.8-fold) were significantly elevated in zebrafish fed the S0.15 diet (p < 0.05; Figure 6A). Of the lipid-degradation-related genes, carnitine palmitoyltransferase 1a (cpt1a), acyl-CoA oxidase 3 (acox3), and enoyl-CoA hydratase and 3-hydroxyacyl CoA dehydrogenase (ehhadh) were not affected by the S0.15 diet compared with the CK diet (p > 0.05; Figure 6B).

In the liver, the expression levels of lipid synthesis-related genes acc1 (1.6-fold), fas (1.8-fold), and diacylglycerol acyltransferase 2 (dgat2; 2.7-fold) were significantly increased by S0.15 feeding (p < 0.05; Figure 6C). The marker for lipid-degradation ehhadh (0.47-fold, p < 0.05) was significantly reduced in zebrafish fed the S0.15 diet, while the other markers tested, cpt1a and acox3 were not regulated (Figure 6D).

In the muscle, lipid synthesis-related genes acc1 (1.9-fold, p < 0.05) and fas (3.2-fold, p < 0.05) were significantly upregulated in fish fed the S0.15 diet while dgat2 was not affected (Figure 6E). The expression levels of lipid-degradation-related genes cpt1a (0.41-fold, p < 0.01) and acox3 (0.2-fold, p < 0.01) were significantly downregulated in zebrafish receiving the S0.15 diet, while ehhadh (0.27-fold, p = 0.07) was not affected although it had a clear tendency of downregulation in the S0.15 group (Figure 6F). These results indicate that dietary succinate might promote lipid anabolism in the intestine, liver, and muscle.

The expression levels of intestinal genes involved in gluconeogenesis glucose-6-phosphatase catalytic subunit 1a (g6pc1a; 0.49-fold) and phosphoenolpyruvate carboxykinase 1 (pck1; 0.45-fold) were significantly downregulated in zebrafish fed the S0.15 diet (p < 0.01; Figure 7A), whereas the expression levels of the genes in the glycolysis and glycogen synthesis pathways, phosphorylase kinase (pk), phosphofructokinase (pfk), and glycogen synthase 2 (gys2), were not affected by the S0.15 diet (Figures 7B,C).

In the liver, the expression levels of the hepatic gluconeogenesis-related genes g6pc1a (2.33-fold) and pck1 (1.44-fold) were significantly upregulated when fed the S0.15 diet (p < 0.05; Figure 7D), while the expression levels of glycolysis-related genes showed no change (Figure 7E). Glycogen synthesis-related gene gys2 was significantly upregulated in the S0.15 group (2.25-fold, p < 0.05; Figure 7F) compared with the CK group.

In the muscle, there were no effects of the S0.15 diet on the expression of gluconeogenesis-related genes (Figure 7G), whereas glycolysis-related genes pfk (1.87-fold, p < 0.05) increased with S0.15 feeding and pk had a non-significant increase as did (Figure 7H). The expression of glycogen synthesis-related gene gys2 was significantly upregulated in the S0.15 group (1.86-fold, p < 0.05; Figure 7I) compared with the CK group. These results indicate that dietary succinate could inhibit intestinal gluconeogenesis but promote hepatic gluconeogenesis. In addition, glucose utilization in the liver and muscle might be increased by succinate feeding.

The expression of intestinal genes related to protein synthesis, mammalian target of rapamycin (mtor), and asparagine synthetase (asns), were not affected by the S0.15 diet (Figure 8A), while protein degradation-related genes glutamate dehydrogenase 1a (gdh1a; 0.54-fold), aminopeptidase N (apn; 0.62-fold) were significantly downregulated by S0.15 feeding (p < 0.01; Figure 8B) compared with CK feeding.

Similarly, the expression of the same genes related to hepatic protein synthesis was not significantly affected by the S0.15 diet (Figure 8C), while S0.15 feeding significantly downregulated protein degradation-related genes gdh1a (0.68-fold) and apn (0.48-fold) compared to CK feeding (p < 0.05; Figure 8D).

The expression levels of the same genes related to muscular protein synthesis were not significantly affected by diets (Figure 8E). As in the intestine and liver, the S0.15 diet appeared to downregulate the gene markers for protein degradation gdh1a (0.67-fold, p < 0.01) and apn (0.41-fold, p < 0.05), as well as a non-significant reduction of pept1 expression in the muscle (Figure 8F). These results suggest that dietary succinate could inhibit protein catabolism of the intestine, liver, and muscle, suggesting protein-sparing effects.

The distribution of intestinal protein succinylation in zebrafish fed the CK or the S0.15 diet was confirmed by western blotting using an anti-succinyllysine antibody (Supplementary Figure 2A). According to the western blotting results, the molecular weight of intestinal succinylated proteins was 25–100 kD (Supplementary Figure 2A). Upon S0.15 feeding, intestinal proteins with a molecular weight of 25–35 kD had increased succinylation levels, while proteins with a molecular weight of 35–70 kD had decreased succinylation levels compared with CK feeding (Supplementary Figure 2A). A total of 6,949 succinylated peptides were identified from the 28,608 MS/MS spectra with a FDR of <1%, and 1,644 succinylated lysine sites were identified in 588 succinylated proteins. The summary of all differentially modified proteins and sites in the intestine is presented in Supplementary Figure 2B. Among the intestinal proteins, succinylation levels of 41 sites of 29 proteins were upregulated, while succinylation levels of 29 sites of 19 proteins were downregulated in the S0.15 group (Supplementary Figure 2B) compared to the CK group.

Western blotting results showed that the molecular weight of hepatic succinylated proteins was 25–130 kD (Supplementary Figure 2C). Upon S0.15 feeding, hepatic proteins with a molecular weight of 25–35 kD had increased succinylation levels, while proteins with a molecular weight of 70–130 kD had decreased succinylation levels compared with CK feeding (Supplementary Figure 2C). A total of 8,751 succinylated peptides were identified from the 36,755 MS/MS spectra with a FDR of <1%, and 1,481 succinylated lysine sites were identified in 483 succinylated proteins. The summary of all differentially modified proteins and sites in the liver is presented in Supplementary Figure 2D. Compared with the CK group, only nine sites of eight proteins in the S0.15 group showed upregulated succinylation levels, while 59 sites of 49 proteins showed downregulated succinylation levels (Supplementary Figure 2D).

In the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, intestinal proteins with upregulated succinylation levels were enriched in oxocarboxylic acid metabolism and TCA cycle (Figure 9A), including isocitrate dehydrogenase 3b (Idh3b), citrate synthase (Cs), glutamic-oxaloacetic transaminase 2b (Got2b), Pck2, and malate dehydrogenase (Mdh2). Proteins with downregulated succinylation levels were part of pathways for fatty acid and amino acid degradation, including hydroxyacyl-CoA dehydrogenase trifunctional multienzyme complex subunit alpha a (Hadhaa), Hadhab, and enoyl-CoA hydratase, short chain 1 (Echs1; Figure 9B).

Among the intestinal proteins with increased succinylation levels, Cs was highly interactive with other proteins, including Mdh2, acyl-CoA dehydrogenase medium chain, Idh3b, Got2b, Pck2, heat shock protein family D member 1 (Hspd1), and ATP synthase alpha subunit (Atp5a1; Supplementary Figure 3A). Similarly, a network of interaction was formed among proteins with reduced succinylation levels, including superoxide dismutase 2, hydroxysteroid 17-beta dehydrogenase 10 (Hsd17b10), Hadhaa, Hadhab, and Echs1 (Supplementary Figure 3B). These results suggest that dietary succinate could negatively or positively influence intestinal protein succinylation levels.

Hepatic proteins with increased succinylation levels were observed in the oxidative phosphorylation and phenylalanine, tyrosine, and tryptophan biosynthesis pathways, including Atp5a1, Atp5c1, Atp5h, and Got2a proteins (Figure 10A). Proteins with decreased succinylation levels were found within protein processing and transport pathways in the endoplasmic reticulum (prolyl 4-hydroxylase, beta polypeptide, Sec61 translocon subunit alpha 1 (Sec61a1), protein disulfide isomerase family A member 4, Hsp90b1, and Hspa5), the primary bile acid biosynthesis pathway (sterol carrier protein 2a and Hsd17b4) and cysteine and methionine metabolism (betaine-homocysteine methyltransferase, adenosylhomocysteinase, and methionine adenosyltransferase 1A; Figure 10B).

An interaction was observed between Atp5a1, Atp5c1, and Atp5h, which had increased succinylation levels (Supplementary Figure 4A). Proteins with decreased succinylation levels formed three interaction niches involving the ribosome [ribosomal protein L22 (Rpl22), Rplp2l, ribosomal protein S25 (Rps25), Rpl25, Rps11, Rps21, Rps3a, Rpl4, Rpl11, and Sec61a1], glucose metabolic process [glyceraldehyde-3-phosphate dehydrogenase (Gapdh), triosephosphate isomerase 1b, phosphoglucomutase 1, aldehyde dehydrogenase 9 family, member A1a, and enolase 3], and protein processing and transport in the endoplasmic reticulum (Supplementary Figure 4B). Different from the case of the intestine, dietary succinate might lead to decreased levels of protein succinylation in the liver.

The α diversity indexes of the gut microbial community, including Shannon, Simpson, and Chao, showed no differences between the S0.15 and the CK group (Table 1). However, PCA analysis indicated a different gut community composition at the OTU level with PC1 variances of 89.09% and PC2 of 8.04% (Figure 11A). At the phylum level, there was a significant increase in the relative abundance of Proteobacteria (60.24 vs. 37.04%, p < 0.01) and a reduction in the relative abundance of Fusobacteria (23.38 vs. 49.75%, p < 0.01) in the S0.15 group compared with the CK group (Table 2 and Figure 11B). Moreover, the ratio of Fusobacteria to Proteobacteria was significantly decreased in S0.15-fed zebrafish (p < 0.05; Table 2). At the genus level, the abundance of Cetobacterium (23.38 vs. 49.75%, p < 0.01) was significantly reduced in the S0.15 group compared to the CK group (Table 3 and Figure 11C). These results suggest that dietary succinate promotes the growth of Proteobacteria but diminishes Fusobacteria and Cetobacterium.